Search results for " Marko"
showing 10 items of 201 documents
Vector representation of non-standard spellings using dynamic time warping and a denoising autoencoder
2017
The presence of non-standard spellings in Twitter causes challenges for many natural language processing tasks. Traditional approaches mainly regard the problem as a translation, spell checking, or speech recognition problem. This paper proposes a method that represents the stochastic relationship between words and their non-standard versions in real vectors. The method uses dynamic time warping to preprocess the non-standard spellings and autoencoder to derive the vector representation. The derived vectors encode word patterns and the Euclidean distance between the vectors represents a distance in the word space that challenges the prevailing edit distance. After training the autoencoder o…
The dynamic interdependence in the demand of primary and emergency secondary care: A hidden Markov approach
2021
This paper develops an extension of the class of finite mixture models for longitudinal count data to the bivariate case by using a trivariate reduction technique and a hidden Markov chain approach. The model allows for disentangling unobservable time-varying heterogeneity from the dynamic effect of utilisation of primary and secondary care and measuring their potential substitution effect. Three points of supports adequately describe the distribution of the latent states suggesting the existence of three profiles of low, medium and high users who shows persistency in their behaviour, but not permanence as some switch to their neighbour's profile.
Space-time analysis of GDP disparities among European regions : a Markov chains approach
2004
The purpose of this paper is to study the evolution of the disparities between 138 European regions over the 1980-1995 period. We characterize the regional per capita GDP cross-sectional distribution by means of nonparametric estimations of density functions and we model the growth process as a first-order stationary Markov chain. Spatial effects are then introduced within the Markov chain framework using regional conditioning (Quah, 1996b) and spatial Markov chains (Rey, 2001). The results of the analysis indicate the persistence of regional disparities, a progressive bias toward a poverty trap and the importance of geography to explain the convergence process.
A study on forecasting electricity production and consumption in smart cities and factories
2019
Abstract The electrical power sector must undergo a thorough metamorphosis to achieve the ambitious targets in greenhouse gas reduction set forth in the Paris Agreement of 2015. Reducing uncertainty about demand and, in case of renewable electricity generation, supply is important for the determination of spot electricity prices. In this work we propose and evaluate a context-based technique to anticipate the electricity production and consumption in buildings. We focus on a household with photovoltaics and energy storage system. We analyze the efficiency of Markov chains, stride predictors and also their combination into a hybrid predictor in modelling the evolution of electricity producti…
Capacity Upper Bound of Channel Assembling in Cognitive Radio Networks with Quasistationary Primary User Activities
2013
In cognitive radio networks (CRNs) with multiple channels, various channel-assembling (ChA) strategies may be applied to secondary users (SUs), resulting in different achieved capacity. However, there is no previous work on determining the capacity upper bound (UB) of ChA for SUs under given system configurations. In this paper, we derive the maximum capacity for CRNs with ChA through Markov chain modeling, considering that primary user (PU) activities are relatively static, compared with SU services. We first deduce a closed-form expression for the maximum capacity in a dynamic ChA strategy and then demonstrate that no other ChA strategy can provide higher capacity than that achieved by th…
Aggregated Packet Transmission in Duty-Cycled WSNs: Modeling and Performance Evaluation
2017
[EN] Duty cycling (DC) is a popular technique for energy conservation in wireless sensor networks (WSNs) that allows nodes to wake up and sleep periodically. Typically, a single-packet transmission (SPT) occurs per cycle, leading to possibly long delay. With aggregated packet transmission (APT), nodes transmit a batch of packets in a single cycle. The potential benefits brought by an APT scheme include shorter delay, higher throughput, and higher energy efficiency. In the literature, different analytical models have been proposed to evaluate the performance of SPT schemes. However, no analytical models for the APT mode on synchronous DC medium access control (MAC) mechanisms exist. In this …
CAD-Based Training of an Expert System and a Hidden Markov Model for Obstacle Detection in an Industrial Robot Environment
2012
Abstract Deploying industrial robots in harsh outdoor environments require additional functionalities not currently provided. For instance, movement of standard industrial robots are pre-programmed to avoid collision. In dynamic and less structured environments, however, the need for online detection and avoidance of unmodelled objects arises. This paper focus on online obstacle detection using a laser sensor by proposing three different approaches, namely a CAD-based Expert System (ES) and two probabilistic methods based on a Hidden Markov Model (HMM) which requires observation based training. In addition, this paper contributes by providing a comparison between the CAD-based ES and the tw…
A structured filter for Markovian switching systems
2014
In this work, a new methodology for the structuring of multiple model estimation schemas is developed. The proposed filter is applied to the estimation and detection of active mode in dynamic systems. The discrete-time Markovian switching systems represented by several linear models, associated with a particular operating mode, are studied. Therefore, the main idea of this work is the subdivision of the models set to some subsets in order to improve the detection and estimation performances. Each subset is associated with sub-estimators based on models of the subset. In order to compute the global estimate and subset probabilities, a global estimator is proposed. Theoretical developments ba…
Uncommon Suffix Tries
2011
Common assumptions on the source producing the words inserted in a suffix trie with $n$ leaves lead to a $\log n$ height and saturation level. We provide an example of a suffix trie whose height increases faster than a power of $n$ and another one whose saturation level is negligible with respect to $\log n$. Both are built from VLMC (Variable Length Markov Chain) probabilistic sources; they are easily extended to families of sources having the same properties. The first example corresponds to a ''logarithmic infinite comb'' and enjoys a non uniform polynomial mixing. The second one corresponds to a ''factorial infinite comb'' for which mixing is uniform and exponential.
Using Hankel matrices for dynamics-based facial emotion recognition and pain detection
2015
This paper proposes a new approach to model the temporal dynamics of a sequence of facial expressions. To this purpose, a sequence of Face Image Descriptors (FID) is regarded as the output of a Linear Time Invariant (LTI) system. The temporal dynamics of such sequence of descriptors are represented by means of a Hankel matrix. The paper presents different strategies to compute dynamics-based representation of a sequence of FID, and reports classification accuracy values of the proposed representations within different standard classification frameworks. The representations have been validated in two very challenging application domains: emotion recognition and pain detection. Experiments on…